Beam-steering the way to real-world mmWave 5G

Telecoms.com periodically invites expert third parties to share their views on the industry’s most pressing issues. In this piece Esat Sibay, COO and CFO at ALCAN Systems looks at how to get the most out of millimetre-wave 5G.

The past 12 months have seen the much-anticipated arrival of 5G but, so far, it’s not quite living up to the associated hype. If we take for example the US market, which is making some of the biggest strides in this field, operators have been firmly focused in low band frequencies. While there is still a great deal of value in this iteration of 5G, even T-Mobile has conceded, low band 5G will only be 20% faster than existing 4G LTE networks. Undoubtedly this is a move forward for the industry but, it is not a network technology that is going to see all the promises of 5G, and the ‘fourth industrial revolution’ that now seems synonymous with it, come to fruition. The truth is, to get even close to realising the potential of 5G, operators must enter the unchartered waters of mmWave 5G.

Why mmWave?

mmWave frequencies have long been associated with 5G delivery, but they are not without drawbacks. There are two big problems we encounter when using mmWave frequencies for 5G; poor penetration – a wall, or even a user’s hand can block signal, and a limited range of only 1000ft; that’s 2% of the range of 4G. So, when it comes to designing a commercially viable 5G network that can deal with these restrictions in a real-world environment, mmWave 5G poses a significant challenge for operators.

With such huge issues to navigate, which fuel uncertainty around its suitability, the question is, why bother with mmWave for 5G? The answer is simple, the 20% improvement on 4G that is possible in the low bands isn’t enough. It is unable to support dozens of proposed 5G use cases and puts a ceiling on the return operators will see on 5G investments. However, overcome the problems associated with mmWave and you are left with a technology capable of delivering above and beyond the promises of 5G. Take Telstra, only this week it has announced mmWave 5G trials that could achieve speeds up to 8 times faster than 4G. High-bandwidth, super-speeds, high availability and a significant improvement on 4G make mmWave frequencies ideal. This is of course is why the industry is so keen to tap into their potential– but how?

The boons of beam-steering

One of the key technologies that allow 5G to reach its potential by overcoming the restrictions of mmWave is beam-steering. As the name suggests, it allows a signal to be focused in a particular direction, rather than radiating 120º as it normally would. The signal, which is controlled with Electronically Steerable Antennas (ESAs) enables precise propagation and a faster and more reliable connection than would otherwise be possible. It minimises penetration losses and increases the reach of 5G working in mmWave frequencies.

In principle this sounds like the perfect answer; simply steering around obstacles to enable 5G, but inevitably there is a catch. The catch is that ESAs that so effectively tackle the challenges of mmWave penetration, are typically too high cost to be a realistic option for most operators. Based on existing ESA technology, operators are looking at tens of thousands of dollars per antenna. This may be palatable if only a few per city were required, but given the shorter reach of mmWave frequencies, even when enhanced by beam-steering, it is a technology that must be deployed in high volumes to be effective. Manipulating radiation patterns to navigate the physical limitations of mmWave frequencies is crucial, but operators need to be able to recoup the costs of equipment. So, the next task they face is finding technology to deliver beam-steering at the scale they need and with costings that allow them to have a commercially viable 5G network.

The answer: you could be looking at it

There are many different ways of approaching this. Looking at military solutions for beam-steering, using repeater antennas to expand coverage, but one approach uses technology you are very literally staring at.

Liquid Crystal, the material that is used in most screens across the globe, can be used as a way of developing ultra-low-cost, energy efficient and low-profile beam-steering antennas that offer a potential cure to many of the headaches associated with mmWave 5G.

First explored as part of a research project at Darmstadt University, Liquid Crystal Antennas take what is now an everyday material and evolve its applications. This results in antennas that allow operators to manipulate radiation patterns like any other beam-steering antenna, but the materials used mean the resulting ESA can be built 10 times cheaper than traditional options. Not only this but by using Liquid Crystal to enable beam-steering, the antenna functions with low-power and can be built with a form factor to minimise visibility – think how thin a typical mobile device can be made and this puts it into perspective.

Most technologies that are being developed are being stretched to their physical limitations to work in mmWave frequencies, however a quirk of using Liquid Crystal as the core material for ESAs is performance improves in higher frequencies. This makes it not only a viable solution to 5G stumbling blocks, but it has the potential to thrive in a 6G world.

Conclusion

There is no two ways about it using mmWave frequencies is the only way operators can build a 5G network that lives up to the promises and hype we have seen over the past few years. That said, delivering in these frequencies is a significant challenge which requires operators to completely rethink the architecture of their Radio Access Networks. It needs equipment that does it all – improved RF performance, with a smaller physical footprint, and lower power consumption, all with a reduced cost of ownership.

Operators across the globe have acknowledged that mmWave frequencies must be a key aspect of 5G networks. However, a big question mark still hangs over the best way to achieve this. Regardless of which technologies operators look to, to ensure 5G is a success, it is critical that in striving to achieve mmWave 5G they do not fall into an investment pit of spiralling infrastructure costs that cannot be recouped.

 

Esat Sibay is CFO and COO at ALCAN Systems, a specialist in the development of smart antennas.  He is responsible for financial management, business development, legal and administration at ALCAN and has more than 20 years of experience in finance and strategic consulting with companies such as HSBC, Citigroup and Accenture. He also holds an M.Sc. in Finance from London Business School, a Diploma in Economics from the London School of Economics and a B.Sc. in Industrial Engineering from Bosphorous University.

Verizon starts toying around with mid-band spectrum

With 5G falling flat in the US, it appears Verizon is taking matters into its own hands with an application to the FCC to experiment with mid-band spectrum, specifically, 3.7-3.8 GHz.

In fairness to the US telcos, there hasn’t been much opportunity to deliver 5G over the airwaves which are proving critical to the rest of the world. The ‘C-band’ spectrum is congested, though the FCC is currently in the process of clearing it and creating a dynamic spectrum sharing initiative which could be the envy of the world. Better late than never.

According to the application made to the FCC, Verizon is planning on running trials over the 3.7-3.8 GHz spectrum in several locations in three states, namely:

  • Basking Ridge, New Jersey
  • Westlake, Texas
  • Williamston, Michigan
  • Okemos, Michigan
  • Jenison, Michigan
  • Hudsonville, Michigan
  • Ada, Michigan
  • Lowell, Michigan
  • Sunnyvale, California

Many telcos around the world have been bragging of the benefits of mid-band spectrum, benefiting from a more palatable compromise between increased download speeds and coverage, the US telcos have been struggling with mmWave or low-band airwaves, neither of which can deliver on the much-hyped 5G promise.

The status quo of disappointment was fine as long as all the telcos are underwhelming, but there has been a recent development which should worry the likes of Verizon and AT&T.

As part of the merger agreement between T-Mobile US and Sprint, the new company will have access to all three tiers of spectrum. T-Mobile had been offering 5G over 600 MHz and mmWave already, which was not satisfactory, however it now has access to Sprint’s 2.5 GHz assets. A blend of low-, mid- and high-band spectrum licences should see a very effective delivery of 5G. This is already being delivered in Philadelphia, though it won’t be long until it is scaled by the ambitious challenger.

Looking at the 5G subscriber forecasts by analyst firm Omdia, this could have a very material impact on the balance of power in the US telco industry.

Forecast of 5G subscriptions in US (2020-2022)
Telco 2020 2021 2022
AT&T 5,581,572 14,416,872 29,301,757
Verizon 2,520,867 16,560,150 35,020,621
T-Mobile and Sprint 5,560,802 18,560,447 36,266,014

Source: Omdia World Information Series

Alone, T-Mobile would erode the subscription lead AT&T and Verizon hold over it today, but it would still be in third place. When you combine the T-Mobile and Sprint figures, you have a market leading firm.

Some might suggest the figures are incorrect as the merger would mean Sprint disappears, but this will not happen overnight. Legacy deals might well be kept in play for the short-term under the Sprint brand as integration projects and campaigns run, but they will be delivered over the same network. The very network which will have the most comprehensive and attractive blend of spectrum.

“Mid-band spectrum provides the sweet spot combination of capacity and coverage for modern 5G networks that the rest of the world is coalescing behind,” Chris Pearson, President of 5G Americas, recently wrote on a blog post championing 5G as a catalyst for recovery from the current global pandemic.

“The international standards forum 3GPP identified the spectrum range 3.3-4.2 GHz as the core 5G band for countries around the world. But the US has yet to auction any exclusive use licensed spectrum in that global mid-band range for 5G.”

Pearson has pointed to regulatory restrictions slowing progress in accessing mid-band spectrum, a critical component in ensuring 5G meets the promises being made by the telecoms industry. A lack of mid-band spectrum is problematic for numerous reasons.

Firstly, coverage can only be delivered only low-band airwaves, but this does not deliver speed upgrades as T-Mobile customers are finding out. Over mmWave means coverage is very limited, which AT&T and Verizon customers are discovering, while it means network deployment is also a lot more expensive as densification projects are very costly and time consuming. Latency is also falling short of all standards by all telcos.

Pearson is of course a champion for the telecoms industry, but the necessity of mid-band spectrum is also replicated at regulatory level.

“For America to be a global leader and win the race to 5G technologies, which we must do for both economic and national security reasons, we must actively identify and make available a key ingredient necessary for 5G networks and systems: mid-band spectrum,” FCC Commissioner Mike O’Reilly said in a letter to President Donald Trump in April.

“Yet, the pipeline is nearly empty, and our wireless providers lack sufficient mid-band spectrum to meet the exponential growth enabled by 5G networks and expected by users. I believe that only you personally, with your unique ability to cut through the bureaucratic stonewalling, can free the necessary spectrum bands to provide our wireless providers the means to succeed.”

If the US is to deliver the 5G promise it needs access to mid-band spectrum. Not only will this benefit consumers, but it will allow enterprise customers to deliver on the newly emerging 5G-powered business models. Without it, US corporations might fall behind international rivals who exist in countries where the mid-band airwaves are available. This is a mid- to long-term consequence, but one which would be much more damaging to the US economy on the whole.

As it stands, only T-Mobile is in an adequate position. This should be a concern for AT&T and Verizon.

T-Mobile is a company which has been very successful in recent years, growing from a position of irrelevance to a genuine threat. The comfortable spectrum position could act as another catalyst for growth, potentially creating a new leader in the US telecoms industry.


Telecoms.com Daily Poll:

How critical is mid-band spectrum in delivering 5G services?

Loading ... Loading ...

What we learned about Dish during the earnings call

With Dish executives leading the company’s quarterly earnings call, details of the plan to crack into the US mobile market were revealed.

The next few years are critical for the US telecoms industry but also the credibility of the FCC and the Department of Justice. Both of these authorities dismissed opposition to the T-Mobile US and Sprint merger, ignoring suggestions it would damage competition. Dish was the reason competition could be maintained, irreversibly changing the US telecoms industry, so it better succeed.

Fortunately, the is being fairly transparent about developments, or certainly more so than most telecoms executives are. But what did we learn from CEO Erik Carlson and Chairman Charlie Ergen last week?

Firstly, $10 billion should be enough to build a nationwide network.

This is a figure which has been banded around quite a lot in recent months without any in-depth explanation, but Ergen believes $10 billion should be enough to meet FCC regulatory requirements and go beyond to create a nationwide network which can compete. There might be a few unforeseen expenditures, spectrum auctions for example, but the team is standing by this estimation.

While the Boost business has not been officially closed yet, the team should have launched in one market by the end of the year, with its own independent core but leaning on the T-Mobile access network. This MVNO agreement will be running for seven years, but the team have already begun talks with tower companies to push forward to create its own network.

What is worth noting is that this work is running independent of the assets which can be purchased from the new T-Mobile company. EVP of Corporate Development Tom Cullen highlighted that deployment planning has begun but once the Boost deal closes, Dish will also have first refusal to acquire cell sites from T-Mobile which are deemed surplus to requirements thanks to the network rationalisation process between T-Mobile US and Sprint.

Although this is detail which some might not have expected, there are still quite a few questions remaining. That said, there is absolute clarity on one area in particular.

“We also took a $356 million impairment charge during the quarter, related primarily to our narrowband IoT build and our satellites D1 and T1,” said CFO Paul Orben. “Now that the T-Mobile/Sprint merger has closed and there is more clarity surrounding our revised build-out requirements. We no longer intend to finish our narrowband IoT build.”

NB-IOT has been struggling to live up to the expectation in numerous markets and this will not help matters. Dish is officially turning its back on NB-IOT, choosing to take an impairment charge on FCC commitments and turn attentions to a 5G network instead of completing the project.

While this might not be the most encouraging of signs, the embracement of OpenRAN and Mavenir as the company’s first official supplier is.

“Marc [Marc Rouanne – Chief Network Officer] continues to work on the architecture and further vendor selection,” said Ergen. “So I would anticipate more of those announcements in the third quarter. And then we’ll share our deployment plans once those are formalized likely on the next call.”

The dynamic of network suppliers is an interesting one for Dish. Ergen highlighted there was a desire to use Huawei equipment, which he described as “best in class”, though the team is being asked to find innovation in new ways. We also found out there is an active dialogue between Dish and Japan’s Rakuten to learn about OpenRAN deployments in the wild.

This is an area many will be keeping a close eye on, not only for validation of a technology which is still not the real deal, but also vendor appointments. The scale of this network, and the aggressive deployment schedule, could force OpenRAN start-ups to grow very quickly. Dish could be a major catalyst for growth for the lucky few who are selected.

It is of course early days, but there are some very interesting developments to keep an eye on here. The team might have opened the door slightly, but there is still much left to discover.

Will the team be able to deploy a network for $10 billion? How will it build its wholesale business unit? When will network slicing begun to be factored in? Which OpenRAN suppliers will be added to the roster over the next few months? Which markets will the postpaid products be launched in first?

With the next earnings call scheduled for July 30, the next three months could offer some very interesting announcements.

Verizon the biggest winner of the latest US millimeter-wave auction

An investment of $1.6 billion got Verizon almost five million licenses in the US auction of the Upper 37 GHz, 39 GHz, and 47 GHz bands that will be used for 5G.

You can see who got the most below. Apparently T-Mobile had been expected, at least by some analysts, to be the big winner, but it ended up a distant third. It’s also worth noting by how much Verizon won the auction, dropping almost half a billion bucks more than second placed AT&T. We’d be lying if we said we knew why there was so much variation in the price per license, but Columbia Capital must have really fancied those 52 it won.

Winner Payment Licenses
Verizon $1,624,101,808 4,940
AT&T $1,185,734,976 3,267
T-Mobile $872,791,192 2,384
Columbia Capital $306,711,619 52
Dish $202,532,574 2,651
U.S. Cellular $146,342,281 237
Sprint $113,948,318 127

“The successful conclusion of Auction 103—the largest amount of spectrum offered in an auction in U.S. history—is one more significant step the FCC has taken toward maintaining American leadership in 5G,” said FCC Chairman Ajit Pai. “A critical part of our 5G FAST plan is pushing more spectrum into the commercial marketplace. Last year, the FCC auctioned the 28 GHz and 24 GHz bands.

“All told, those two auctions and this one have made available almost five gigahertz of high-band spectrum for commercial use. To put that in perspective, that is more spectrum than is currently used for terrestrial mobile broadband by all wireless service providers in the United States combined. Auction 103 was a tremendous success, and we look forward to building on this positive result with the 3.5 GHz auction, which is scheduled to begin on June 25, and the C-band auction, which is scheduled to begin on December 8.”

Have you noticed how much Americans like the word ‘tremendous’ these days? One definition of it is ‘being such as may excite trembling or arouse dread, awe, or terror’. Fair enough. The mid bands on offer later this year are also for 5G. Those frequencies have better propagation characteristics than millimeter wave but there will be less of them, as is so often the way with radio spectra.

TIM claims 5G speed record with help from Ericsson and Qualcomm

A trio of telecoms trailblazers managed to break the 2 Gbps barrier with 5G over 26 GHz, which is apparently a European record.

TIM, Ericsson and Qualcomm have a rich history of collaborating in the name of self-promotion over 5G and there’s nothing like a speed record for a bit of corporate chest-beating. TIM has a whopping 400MHz of spectrum in this millimetre wave band, which is the main reason it’s able to set records such as this. Meanwhile, as ever, Ericsson provided the radio and Qualcomm the modem.

“This milestone paves the way to the development of new 5G solutions to grant fixed ultrabroadband to families, companies and public authorities not yet covered,” said Michele Gamberini, TIM’s CTIO. “This also includes coverage dedicated to the development of robotics and automation digital services in the smart manufacturing area. All of our customers will therefore be able to take advantage of a wide range of integrated solutions that will allow them to fully enter the digital society”.

“We are extremely pleased that TIM has chosen Ericsson’s 5G technology to achieve this important milestone, placing our country at the forefront of the commercial implementation of the fifth generation of mobile networks,” said Emanuele Iannetti, Country Manager at Ericsson Italy. “Ericsson thus confirms its technological leadership and its readiness to anticipate any market demands.”

“Qualcomm Technologies congratulates TIM on this significant milestone which again demonstrates the potential of 5G mmWave technology and shows how operators are able to use a wide range of spectrum bands to deploy 5G,” said Enrico Salvatori, president, Qualcomm EMEA. “2020 will see a significant expansion in 5G coverage and the use of mmWave bands will play a clear role in the build-out.”

The rest of the TIM press release was mostly spent going on about how this proves the company was right to blow loads of cash at the last Italian spectrum auction. It still remains to be seen how useful high frequency spectrum will be in real life, or indeed how much use there will be for such high data rates, but it’s always nice to be able to claim you’re at the cutting edge regardless.

IRU WRC-19 earmarks some more high frequency spectrum for 5G

After a month of working on their tans in the Egyptian resort of Sharm El-Sheikh, the global telecoms community has identified a bunch more spectrum it wants to dedicate to 5G.

The latest ITU World Radiocommunication Conference inevitably had 5G as a major theme. The point of the event is to get global consensus on certain telecoms matters and spectrum harmonisation if often high on the list. If different parts of the world use different spectrum for mobile then it makes roaming a massive hassle and drives up the cost of components, so it’s important to get global agreement on this stuff.

So one of the most significant developments was the identification of s bunch more high frequency spectrum that everyone agreed should be set aside for 5G use. Specifically we’re talking about the 24.25-27.5 GHz, 37-43.5 GHz, 45.5-47 GHz, 47.2-48.2 and 66-71 GHz bands.

“The hard won agreements at WRC-19 will favourably impact the lives of billions of people around the world, creating a digital landscape for sustainable growth and development,” said Mr Mario Maniewicz, Director of the ITU Radiocommunication Bureau. “The achievements of WRC-19 in enabling new communication technologies and the protection of existing services will be reflected in the continuous growth of the trillion dollar telecommunication and ICT industry.”

“WRC-19 has brought the mobile industry a step closer to making the full power of 5G something everyone can experience,” said Mats Granryd, Director General, GSMA. “Countries struck the right balance in opening up groundbreaking possibilities for 5G while protecting existing radio services worldwide. The mobile industry’s goal going into WRC-19 was to identify enough 5G spectrum to deliver long-lasting socio-economic benefits. WRC-19 delivered on this goal, and also secured a pathway to 5G’s future success in the agenda for WRC-23.”

“Spectrum regulators have recognized that low, mid and high-band frequencies need to be available and harmonised globally as much as is possible to ensure mobile broadband can meet the expanding services both industry and consumers are expecting in the coming years,” said Joe Barrett, President of GSA. “Today’s decision is a significant and welcome step in ensuring the transformational power of 5G can be fully realised in every market around the world.”

How much use all this very high frequency will be remains to be seen. You have to almost be sat on top of the base station to receive it and the global rollout of the zillions of small cells that would be required to make use of it is not off to a flying start. Still it’s nice to know that when we do eventually find an efficient way to make use of it there will be plenty of spectrum to go round.

GSMA squares up to the space industry

The GSMA has released a statement directed at the lobby groups who are attempting to limit access to the valuable mmWave spectrum frequencies over fears it would interfere with weather forecasting.

While the mmWave spectrum has long been heralded as the holy-grail for telcos when attempting to increase download speeds in the 5G era, the space and satellite industry has been attempting to limit access due to interference with various systems including weather forecasting.

No decisions have been formally made, though the GSMA naturally wants to pressure governments into releasing more spectrum as it performs it duties as the industry’s lobby group.

“We can’t let misinformation and the overly protectionist attitudes of the space industry derail the 5G revolution,” says Brett Tarnutzer, Head of Spectrum, GSMA.

“Over-stringent protection will limit the spectrum needed for 5G and have huge consequences for society. This could put the economic and innovation bonanza accompanying ultra-fast networks on hold for a generation.”

The GSMA is being fairly obvious with its message here. Ignore the fears of the space industry and give the telcos more spectrum. You shouldn’t really expect anything less from the lobby group either; telcos are screaming out for more of the valuable resource.

This spat dates back to objections from the US National Oceanic and Atmospheric Administration (NOAA) and National Aeronautics and Space Administration (NASA), which unveiled a report suggesting the high-frequency spectrum bands would interfere with weather forecasting systems, potentially decreasing the accuracy. Democrat Senators Ron Wyden (Oregon) and Maria Cantwell (Washington) jumped on a report produced by NOAA and NASA, writing to the Oval Office suggesting a halt on spectrum usage in the 24 GHz bands.

Over the next couple of weeks, it would not be a surprise to see this conflict enter into the back and forth as the World Radiocommunication Conference 2019 is set to start in 100 days-time. At this event, 3000 delegates will attempt to agree on how radio-wave capacity will be used.

The question these delegates will have to answer is what is more important. The space agency has defended the use of the spectrum for weather forecasting, demonstrating its important to safety and various different industries, but the GSMA has done the same. Most governments are looking towards technology and connectivity as a means to generate economic momentum and the swift implementation of 5G is critical to ensure individual nations do not fall behind the global leaders.

What we suspect will happen is a middle-ground will be found, an attempt to appease all parties involved, though no-one is entirely satisfied. This is generally how such bureaucratic exercises tend to unravel.

Money is piling up in the US 24 GHz auction

Over 30 companies have put more than $560 million in bid money on the table at FCC’s auction for the 24 GHz frequency. And this is only the beginning.

Following the underwhelming auction of the 28 GHz (dubbed Auction 101) spectrum, which only returned $703 million, the new auction of the 24 GHz (dubbed Auction 102) is heating up quickly. The auction started last Thursday and has gone through 11 rounds of the first phase of the auction, or the “clock phase”, when participants bid on a Partial Economic Area (PEA) blocks. By the end of round 11, the gross proceeds have reached a total amount of $563,427,235. There are still two days, or six more rounds to go, before the winners can move to the next phase of the process.

The “assignment phase” will allow the winners from the first phase to bid for specific frequency licence assignments. The total bid value for the 24 GHz frequencies could go up to between $2.4 billion and $5.6 billion, according to the estimate by Brian Goemmer, founder of the spectrum-tracking company AllNet Insights & Analytics, when he spoke to our sister publication Light Reading.

The key difference the has driven up the interest from the bidders for Auction 102 is the locations where the frequencies are made available. While major metropolises like New York, Los Angeles, or Chicago, were absent from 28 GHz auction, they are all on the current 24 GHz auction together with other major cities that would be the candidates for the 5G services to roll out in the first wave.

Bidders have included AT&T, Verizon, T-Mobile, Sprint and more than 30 other companies. The FCC will announce the winners including those from Auction 101 only after both phases of Auction 102 are completed.

In addition to bidding for mmWave frequencies, operators like AT&T are also actively refarming the lower frequency bands in their possession that are used to provide 3G services. AT&T sent a notice to its customers in February that it will stop 3G only SIM activation, urging customers to move to LTE. The company said “we currently plan to end service on our 3G wireless networks in February 2022.” Specifically the company is planning to refarm the 850 MHz and 1900 MHz frequency bands, saying “it may be necessary for us to turn down one band of our owned and operated 3G network, such as 1900 MHz or 850 MHz service”.

Considering the AT&T only switched its 2G networks off at the beginning of 2017, this is a clear sign that the generational transition of mobile telecom services is accelerating. Earlier in the middle of last year, Verizon confirmed that it will shut down its 3G CDMA networks by the end of 2019. Even earlier at the MWC in 2017, T-Mobile’s CTO Neville Ray said the company was looking to sunset both GSM and WCDMA.

Qualcomm has one last 2018 5G fest with TIM in Rome

Qualcomm’s events team has had a frantic end to the year, culminating in the demo of a 5G NR video call over millimeter wave in Italy.

The venue was TIM’s freshly unveiled 5G Innovation Hub in Rome. In attendance were Qualcomm and TIM, of course, but also a host of kit and device partners as well as the great and good of Roman public life, including its Mayor Virginia Raggi (pictured). Most of the presentations were in Italian, but they sounded pretty cool, and there were also a bunch of demos from the various partners.

The highlight of the day was what was claimed to be Europe’s first 5G NR video call, completed over the TIM network using millimeter wave. It made use of a Qualcomm modem and some Ericsson kit. The demo is being positioned as ‘a new milestone that will soon lead to the commercial use of 5G mmWave technology in Europe.’

“When we started to define the strategy and the development plans for 5G, we immediately realized that such a massive challenge could not be faced without the support of a wide range of partners committed to the same goal,” said Mario Di Mauro, Chief Strategy, Innovation & Customer Experience Officer at TIM.

“We therefore proposed Qualcomm Technologies set up a place where work on the new 5G services and every business idea could find a quick realization thanks to the support of leading international technology players, innovative partners and start-ups from the local and national ecosystem.”

“Qualcomm Technologies is very excited to be part of this initiative and we would like to congratulate TIM on the significant momentum they have achieved in a short time with the Hub,” said Enrico Salvatori, President Qualcomm EMEA. “A great example of innovation is today’s demo showing the first 5G mmWave mobile smartphone form-factor mobile test device powered by the Snapdragon X50 5G modem connecting to Ericsson 5G Radio Access Network.

“We are very pleased to be part of the team helping to bring 5G to commercial reality in Italy in 2019 and also to realizing the vision of the Hub. 5G is so much more than new devices and smartphones and it will provide significant growth opportunities in new sectors. The Hub provides TIM with a strong platform to leverage the benefits of 5G to a whole host of new customers and industries.”

We’ll leave it at that for now, but we shot a bunch of video interviews while we were there so keep an eye out for those in the coming days. We can also recommend the Farina Kitchen pizza restaurant, which features a proper wood fired oven and does a very naughty fried pizza starter. Here’s a shot of the 5G call taking place.

TIM video call

Going under the hood of Qualcomm Snapdragon 855: plenty to like

More details of Qualcomm’s first 5G chipset have been released, bringing all-round improvements, and a 5G chipset for PCs was also announced.

On the first day of its annual Snapdragon Technology Summit, Qualcomm announced its 5G chipset for mobile devices, the Snapdragon 855, but released limited specs. On the following two days more details were disclosed. An SoC for 5G-connected PCs, the Snapdragon 8cx was also unveiled.

In addition to the X50 modem for 5G connectivity (on both mmWave and sub-6GHz frequencies) and X24 modem (to provide LTE connectivity), at the centre of the Snapdragon 855 is ARM’s new flagship Cortex A76 CPU, marketed by Qualcomm as Kryo 485. It contains 8 cores with the single core top performance at 2.84 GHz. Qualcomm claims the 855 is 45% faster than its predecessor 845, though it did not specify what exactly this refers to. More importantly for Qualcomm, the top speed is 9% faster than the Kirin 980 from HiSilicon (a Huawei subsidiary), another 7-nanometre implementation of the ARM Cortex A76.

Also included in the 855 is the new Adreno 640 GPU rendering graphics. Qualcomm has focused 855’s marketing messages on gaming performance, and the GPU is at the core to deliver it. Qualcomm claims the new GPU will enable true HDR gaming, as well as support the HDR10+ and Dolby Vision formats. Together with the display IP, the Adreno 640 GPU will support 120fps gaming as well as smooth 8K 360-degree video playback. Another feature highlighted is the support for Physically Based Rendering in graphics, which will help improve VR and AR experience, including more accurate lighting physics and material interactions, for example more life-like surface texture, or material-on-material audio interaction.

The key new feature on Snapdragon’s Hexagon 690 DSP is that it now includes a dedicated Machine Learning (ML) inferencing engine in the new “tensor accelerator”. The Hexagon 690 also doubles the number of HVX vector pipelines over its predecessors the Hexagon 680 and 685, to include four 1024b vector pipelines. The doubled computing power and the dedicated ML engine combined are expected to improve the Snapdragon 855’s AI capability by a big margin.

The integrated new Spectra 380 image signalling processor (ISP) will both improve the Snapdragon’s capability to deepen acceleration and to save power consumption when processing images. Qualcomm believes the new ISP will only consume a quarter of the power as its predecessor for image object classification, object segmentation, depth sensing (at 60 FPS), augmented reality body tracking, and image stabilisation.

On the OEM collaboration side, in addition to Samsung, on day 2 of the event we also saw Pete Lau, the CEO of Chinese smartphone maker OnePlus come to the stage to endorse the new 5G chipset and vow to be the “first to feature” the Snapdragon 855. Separately, the British mobile operator EE announced that it will range a OnePlus 5G smartphone in the first half of 2019.

On the same day, thousands of miles away, more Chinese smartphone OEMs including Xiaomi, OPPO, Vivo, and ZTE (in addition to OnePlus) also embraced the new Snapdragon chipset at the China Mobile Global Partner Conference in Guangzhou, southern China. China Mobile will also launch a customer premise equipment (CPE), likely a fixed wireless access modem, using the same platform.

Back in Hawaii, on day 3 of the Snapdragon Tech Summit, Qualcomm launched a new chipset for PC: the Snapdragon 8cx (“c” for computer, “x” for eXtreme). This is Qualcomm’s third iteration of chipset for PC, built on ARM v8.1 (a variant of Cortex A76). Similar to the Snapdragon 855, the 8cx also has the X24 integrated cellular modem with for LTE connectivity, and the X50 modem with 5G connectivity can be paired with it. The CPU also has eight cores, with a top speed of 2.75 GHz. The new Adreno 680 GPU is said to process graphics twice as fast as the GPU in the previous generation ARM for Windows chipset (Snapdragon 850) but 60% more efficient in power consumption.

Perhaps the most meaningful change is its memory architecture. The Snapdragon 8cx will have a 128-bit wide interface, enabling it to provide native support for much more software and applications, including Windows 10 Enterprise and Office 365, which clearly is a sales pitch to the corporate IT departments.

Unlike the OEM support garnered by Snapdragon 855, there was no public endorsement by PC makers yet. Lenovo did come to the stage but was only talking about its Yoga 2-in-1 notebooks that have used earlier generations of Snapdragon chipsets for Windows on ARM. On the other hand, Qualcomm does not position Snapdragon 8cx as a replacement for the 850 but rather as a higher end contemporary, with 850 mainly targeted at a niche consumer market.

In general, this year’s Snapdragon Tech Summit has delivered more step change with the new product launches. More concrete industry support was also on show, indicating that, depending on how fast and extensive 5G is to be rolled out, we may start seeing true 5G smartphones in the first half of next year. We may need to wait a bit longer before a reasonable line-up of always-on 5G connected PCs can hit the market.